Differential Reporting Errors in Birth Histories by Socioeconomic Group

Yoonjoung Choi, Johns Hopkins University
May Sudhinaraset, Johns Hopkins University

This papers examines differential patterns of reporting errors in birth histories by socioeconomic background, using Demographic and Health Surveys. The birth history data have provided information on childhood mortality in many developing countries where complete vital registration systems lack. The data also have been critical to study inequalities in mortality. Nevertheless, few examined differential data quality by socioeconomic background, which potentially can bias measurements of inequalities in age-specific mortality as well as variation in age-patterns of mortality. We purposely selected 34 Demographic and Health Surveys and calculated indicators of two major reporting errors: digit preference in reporting age at deaths and omission of live births. We assessed differential levels of indicators across household wealth quintiles as well as maternal educational classifications. Findings from this study will be valuable to better understand measurements of socioeconomic variation in levels and age-patterns of childhood mortality in developing countries.